Immunohistochemical studies demonstrate the presence of classical pathway complement component in association with amyloid plaques and neurofibrillary tangles in the brains of patients with AD. Analysis of the specific fragments present suggests complement activation, rather than passive binding of complement proteins to brain tissue. The membrane attack complex (C5b-9) can be found in association with dystrophic neurites and neurofibrillary tangles, suggesting that complement-mediated neuronal lysis may occur. To further investigate whether complement activation occurs in AD, we propose a pilot study of plasma C3a levels in AD patients compared to age- matched controls. C3a is a biologically active fragment released during complement activation. In recent years, measurements of blood levels of complement split products have been shown to be more accurate indicators of complement-mediated disease activity than levels of native complement components (e.g. C3, C4) in diseases such as systemic lupus erythematosus. If C3a levels are elevated in patients with AD, measurement of plasma C3a may be useful in the design and conduct of experimental trials of immunosuppressive therapy. Fifty patients meeting NINCDS-ADRDA criteria for probable AD, fifty non- demented age-matched first-degree relatives of such patients, and fifty age-matched normal controls will be recruited from the ADRC screening programs. Plasma levels of C3a will be measured by a recently developed enzyme immunoassay, and mean levels for the three groups will be compared. The relationship of C3a levels to clinical factors, including age and cognitive function, will be analyzed.